4 papers
Geometric Factorization of Sufficient Harmonic Representations
Kennon Stewart
For tasks of likelihood families invariant under the action of a lie group, the quotient is the minimal sufficient invariant representation. On compact homogeneous spaces, this quo…
Form and Function: Machine Unlearning as a Problem of Misaligned States
Kennon Stewart
We formulate machine unlearning for online L-BFGS as a counterfactual state-alignment problem. Given an actual event stream and a deletion-edited counterfactual stream, the target…
Shape of Memory: a Geometric Analysis of Machine Unlearning in Second-Order Optimizers
Kennon Stewart
We argue that current definitions of machine unlearning are underspecified for second-order optimizers. We compare first-order and second-order learners for their ability to handle…
Mo' Memory, Mo' Problems: Stream-Native Machine Unlearning
Kennon Stewart
Machine unlearning work assumes a static, i.i.d training environment that doesn't truly exist. Modern ML pipelines need to learn, unlearn, and predict continuously on production st…